Future Challenges in Artificial Intelligence for Smart Healthcare

This chapter outlines the major challenges for artificial intelligence (AI) adoption in various aspects of a smart healthcare system. While AI techniques have shown promising improvement in disease diagnosis and prognosis, security remains a major challenge. AI-enabled multiomics analysis helps us to identify new biomarkers and design targeted therapy. The Electronic Health Record (EHR) is a widely adopted mechanism for managing patient information and administering treatment/medicine. Traditionally, labor intensive and expert dependent phenotyping has been used to make them useful for further exploitation as it is really challenging to produce workable analytic models for EHR data for further processing. The paradigm of explainable AI in healthcare is enormous because of its multidimensional perspectives. It involves not only the technical work but also medical, ethical, legal, and social aspects that indeed need in-depth investigation. With more powerful hardware processing, software agility, networking and communication capabilities, AI is becoming a dominant player in smart healthcare systems.